Current Proceedings on Technology

Yazarlar: Murat Topaloglu, Egemen Tekkanat

Konular:-

Anahtar Kelimeler:Mahout,Data leak prevention,Big data,Data security,Machine learning

Özet: Corporal network control has always been a significant problem. The risk that the employers may leak the data that are of corporate importance is a highly substantial security problem. For that reason, the need for systems in data security developed for the reporting, content examination, filtering and categorization of sensitive data becomes more obvious. Text categorization aims to block, report and learn the data that are of corporal importance to pass out via unauthorized means. The results obtained via the text categorization process should be added to the real time DLP architecture immediately. The machine learning methods should be used in order to conduct the text categorization process to be added to the DLP architecture. The interface integrated to the Mahout will be used for the analysis of real time data sets and the network based architecture developed for the DLP system. The comparative experimental results of the data categorization methods were prepared. The selection of categorization method to be used for the momentary control of the data that are of importance is provided.Analyzing the mobile data (data in motion) of the big data sets helps to determine the method of categorization by in the developed architecture. The method preferred for text categorization is applied with Mahout and the analysis of sensitive data and privacy environment is provided.

---